Environmental and Soil Data - Input Data

4. Case Study Description

4.3 Input Data

4.3.2 Environmental and Soil Data

The Table 4.4 and 4.5 gives the data regarding environmental and soil conditions respectively:

Table 4.4: Environmental Data

Parameter Value Unit

Seawater density 1027

Water depth 100 m

Table 4.5: Environmental Data (Soil Conditions)

Parameter Value

Axial friction

LB 0.3

BE 0.5

UB 0.7

Lateral friction 0.8

Soil mobilization 2mm ~ 4mm

With regards to the pipe-soil interaction, it contains a hardly predictable uncertainty in the design of pipeline walking. [18] Hence, the performance of the sensitivity analysis is required to determine the effect of friction on pipeline walking. In the thesis work, the axial friction factor of 0.3 is initially used for the analysis and then other values of factors are applied. Furthermore, the value of 2.0 as the friction factor is additionally used for a wider analysis in the effect of friction factor on the walking phenomenon.

Se-Hoon Yoon 39 | P a g e

0 500 1000 1500 2000

Temp_Load Step 1

5. Results and Discussion

5.1 General

The objective of this chapter is to explicate the results of the thesis work. The methodology described in Ch. 3 is conducted to analyze the pipeline walking phenomenon under the transient thermal cycling. Basically, it presents the responses of the pipeline resting on the seabed described in the analytical FE modeling. The sensitivity study is also discussed in the chapter to understand the critical parameter for the walking phenomenon. Hence, the results of different axial friction factor applications are presented. Lastly, the chapter deals with the necessity of mitigation measures in order to eliminate or prevent pipeline walking in connection with the sensitivity study results. Moreover, the numerical calculation of the walking rate according to SAFEBUCK JIP pipeline design guidelines [15] is separately presented in Appendix I to confirm the validation of the FEA results in the chapter.

5.2 Pipeline Response Analysis Results

This section preferentially deals with the results based on the model conditions as follows:

- Pipe wall thickness: 25.4mm;

- Operating temperature: max. 88 ; - Axial friction factor: 0.3 (LB).

The interpolated temperature load profile is shown as Figure 5.1:

Figure 5.1: Interpolated Temperature Load Profile for 88 KP (m)

Se-Hoon Yoon 40 | P a g e

0 500 1000 1500 2000

load1

5.2.1 Effective Axial Force Profile

The Figures 5.2 through 5.6 show the profiles of the effective axial force developed from the first to the fifth cycle respectively. The “load1” in the Figure 5.2 indicates the initial condition in the pipeline (referring to staying in an ambient temperature). While, it implies restarting the operation in Figures 5.3 through 5.6, which means the first temperature load step is applied to the pipeline in the FEA. The consecutive load numbers stand for the temperature load step application in the FEA. The “load12” displays the shut-down condition in the first cycle, whereas the “load11” refers to that condition in the rest of cycles. Both load numbers are referring to the pipeline turned to the ambient temperature as shutting-down.

Figure 5.2: Effective Axial Force for 1st Cycle

Figure 5.3: Effective Axial Force for 2nd Cycle -800000

0 500 1000 1500 2000

load1

Figure 5.4: Effective Axial Force for 3rd Cycle

Figure 5.5: Effective Axial Force for 4th Cycle

The Figures 5.2 through 5.6 also indicate that as the pipeline is gradually heated, the effective axial force (compressive force) builds up until the pipeline is entirely mobilized (it happens after the 7th load step in this case). Once the pipeline turns to fully mobilized, the profiles show that there are less variations in force over the length of the pipeline. Moreover, The VAP (virtual anchor point) is situated in the middle of the pipeline when fully mobilized. On

0 500 1000 1500 2000

load1

0 500 1000 1500 2000

load1

shut-down condition, which is cooled down at a uniform rate, the contraction takes placed about the VAP.

Figure 5.6: Effective Axial Force for 5th Cycle

The case in the study has the similar aspect of the pipeline behavior that has been presented in Ch.2.4.3.1. The EAF profile is developed in a different form from the second cycle compared with the first cycle profile. It is because that the occurrence of residual tension (caused by friction) when cooling-down. Hence, the pipeline does not return to its original position/size.

According to the Figure 5.3 (2nd cycle), two VAPs are observed when the “load1” is applied in the pipeline, and both are located at KP 28 and KP 1015 respectively. As the pipeline gets heated up, the positions of the VAPs move towards the mid-point and the cold-end (here, at KP 1000 and KP 2000 respectively), specifically it gets the maximum effective axial force at the mid-point (-5.898x N in the case). Other profiles also show the similar aspects, as illustrated in Figures 5.4 through 5.6. Due to that phenomenon, i.e. the shift of the VAP, the asymmetrical expansion from the second heat-up takes place, and the pipeline system in the case study can be expected to “walk”.

-800000 -600000 -400000 -200000 0 200000 400000 600000 800000

0 500 1000 1500 2000

load1 load2 load3 load4 load5 load6 load7 load8 load9 load10 load11

Effective axial force (N)

KP (m) (Figur

Se-Hoon Yoon 43 | P a g e University of Stavanger

5.2.2 Pipeline Cumulative Displacement

The FEA of the case study presents the occurrence of expansion along the full length of the pipeline. The following Figures 5.7 through 5.11 show the profiles of the pipeline expansion developed from the first cycle to the fifth cycle respectively. Similar to the EAF case, the load numbers in the Figures also imply the temperature load steps and operating conditions in the FEA.

Figure 5.7: Pipeline Displacement for 1st Cycle

Figure 5.8: Pipeline Displacement for 2nd Cycle -1

0 500 1000 1500 2000

load1

0 500 1000 1500 2000

load1

Figure 5.9: Pipeline Displacement for 3rd Cycle

Figure 5.10: Pipeline Displacement for 4th Cycle -1

0 500 1000 1500 2000

load1

0 500 1000 1500 2000

load1

Figure 5.11: Pipeline Displacement for 5th Cycle

As the pipeline is progressively hotter, the figures indicate that the non-uniform expansions take place. When KP is adjacent to zero, i.e. close to the hot end, the pipeline is prone to expand towards to its end. Other parts of the line, on the other hand, the shift is made towards the cold end, i.e. KP close to 2000. The figures in Table 5.1 and 5.2 also show displacement in meter at hot/cold ends of the pipeline in each load step.

Table 5.1: Hot-End Pipe Displacement

1st Cycle 2nd Cycle 3rd Cycle 4th Cycle 5th Cycle

0 500 1000 1500 2000

load1

Table 5.2: Cold-End Pipe Displacement

According to Figures 5.7 through 5.11, the expansion at the cold end (i.e. KP 2000) starts to sharply increase with the continuous increase in temperature once the pipeline becomes fully mobilized (in this case, it occurs after “load7”). With regards to the fully mobilized status, it can also be recognized by the values of the pipeline mid-point (i.e. KP 1000) displacement as given in Table 5.3. It shows that less variations of the expansion after the 7th load step in each cycle.

Table 5.3: Mid-Point Pipe Displacement

1st Cycle 2nd Cycle 3rd Cycle 4th Cycle 5th Cycle

5.2.3 Axial Displacement: Walking

The cumulative axial displacement is to be considered to perceive the effect of the thermal cyclic loading on the movement of the pipeline. [1] This section handles axial walking results at the mid-point and two ends due to the expansion and contraction with the subject temperature profiles.

5.2.3.1 Walking at Mid-Point

The Table 5.3 in Ch. 5.2.2 shows the axial displacement at the mid-point of the pipeline through entire cycles (5-cycle). It also presents that each cycle consists of 11 load steps except for the first cycle which has 12 load steps due to the initial condition. Hence, totally 56 temperature load steps are applied to the pipeline model in the FEA. Eventually, the graph of the mid-point displacement under the 5-cycle operation is illustrated as Figure 5.12:

Figure 5.12: Axial Walking at Mid-Point

The Figure 5.12 indicates the walking begins to occur during the second heat-up and becomes stabilized as the operating cycles continue. It is because the constant growth of the displacement under the FE model condition. It shows approximate 0.05m walk in every cycle in the case. The walking rate screening calcuation based on ref. [15] also shows about 0.05m per cycle. Its numerical calculation sheet is presented in Appendix I.

Therefore, it can be assumed that it will have the same displacement increase (i.e. same walking rate) in subsequent cycles after the fifth. If the allowable walking distance is 1.0 meter, it can say that the subject pipeline system will be able to operate 20 startup/shutdown cycles

Number of Operating Cycle (Fi Figure 2.0: Conventional

S-lay installation (Olav Fyrileiv et al., 2005) [9]

Displacement (m)

5.2.3.2 Walking at Two Ends (Hot/Cold)

The data based on the Table 5.1 and 5.2 in Ch. 5.2.2 is used to present the graph of the axial displacement at the two ends (Hot/Cold) of the pipeline over 5 cycles. The temperature load conditions are the same as the mid-point (totally 56 temperature load steps).Accordingly, both ends displacement are illustrated as Figure 5.13 and 5.14 respectively:

Figure 5.13: Axial Walking at Hot End

Figure 5.14: Axial Walking at Cold End -1,2

Number of Operating Cycle (F Figure 2.0: Conventional

S-lay installation (Olav Fyrileiv et al., 2005) [9]

Finally, graphs of axial walking results can be plotted together and it is shown as Figure 5.15:

Figure 5.15: Axial Walking Displacement

The Figure 5.15 shows that over the shut-in and restart cycles, i.e. operations after completing the first cycle, the pipeline has a tendency not to return to the same position/condition. It means that a relative axial movement is made with those repeated cycles and consequently, the pipeline in the case study experiences an accumulated walking.

-1 -0,8 -0,6 -0,4 -0,2 0 0,2 0,4 0,6 0,8 1 1,2

0 10 20 30 40 50 60

Mid-Pint Cold End Hot End

1 ( 0

5 (Figure 2.0:

4 (Figur 3

(Figure 2

Number of Operating Cycle (Fi Figure 2.0: Conventional

S-lay installation (Olav Fyrileiv et al., 2005) [9]

Displacement (m)

Se-Hoon Yoon 50 | P a g e

0 500 1000 1500 2000

load1

5.3 Pipeline Response upon Axial Friction Factor

This section discusses the analysis results regarding the impact of axial friction factors on the walking phenomenon. Basically, three models are generated with different friction coefficients described in Ch.4.3.2, and the additional coefficient value of 2.0 is modeled for the extensive analysis on the walking.

5.3.1 Effective Axial Force Profile

The graphs of the first cycle effective axial force for each model (LB: 0.3, BE: 0.5, UB:

0.7) are illustrated as Figures 5.16 through 5.18 respectively:

Figure 5.16: Effective Axial Force for 1st Cycle with Friction factor 0.3

Figure 5.17: Effective Axial Force for 1st Cycle with Friction factor 0.5 KP (m)

0 500 1000 1500 2000

load1

Se-Hoon Yoon 51 | P a g e

0 500 1000 1500 2000

load1

Figure 5.18: Effective Axial Force for 1st Cycle with Friction factor 0.7

It shows that the EAF increases as the friction factor rises. The VAPs are located in the middle of the pipeline when the pipeline becomes fully mobilized (after “load7” in all cases), and the EAFs at the positions are -5.8x N, -9.7x N and -1.3x N respectively. It is because that the axial resistance (force) per unit length is directly related with the submerged pipeline weight and the friction factor, which is also presented in Ch.2.2.2.2.

However, it gives the different results when the axial friction coefficient becomes bigger than aforementioned factors and reaches a certain value (2.0 in the study). The EAF decreases with that value shown as Figure 5.19. Besides, the force profile is similar to the case of the cyclically constrained pipeline that is presented in Ch.2.3.3.

Figure 5.19: Effective Axial Force for 1st Cycle with Friction factor 2.0 KP (m)

0 500 1000 1500 2000

load1

-1000000

0 500 1000 1500 2000

load1

0 500 1000 1500 2000

load1

Another observation considering variations in axial friction factor is given from the developed EAF profiles during subsequent operating cycles. The force profiles particularly for 4th and 5th cycles under the friction factor 0.5 and 0.7 are shown as Figures 5.20 through 5.23 respectively:

Figure 5.20: Effective Axial Force for 4th Cycle with Friction Factor 0.5

Figure 5.21: Effective Axial Force for 5th Cycle with Friction Factor 0.5

The calculated force data and its profile show that the EAF goes constant (less variation) after

-1500000

0 500 1000 1500 2000

load1

0 500 1000 1500 2000

load1 profiles of 2.0 case are presented separately in Appendix V.

Figure 5.22: Effective Axial Force for 4th Cycle with Friction Factor 0.7

Figure 5.23: Effective Axial Force for 5th Cycle with Friction Factor 0.7

5.3.2 Axial Displacement

This section discusses the axial movements over a range of operating cycles (5-cycle) under different friction coefficient conditions. The results of the displacement in the section mainly deal with the axial displacements at Mid-point, Hot end and Cold end of the pipeline.

5.3.2.1 Mid-Point Axial Displacement

The table 5.4 and 5.5 show the axial displacement at the mid-point of the pipeline with friction factors: 0.5 and 0.7 respectively. Regarding the factor of 0.3 case is presented in Table 5.3 in Ch. 5.2.3.

Table 5.4: Mid-Point Axial Displacement with Friction Factor 0.5

1st Cycle 2nd Cycle 3rd Cycle 4th Cycle 5th Cycle

-0,05

Table 5.5: Mid-Point Axial Displacement with Friction Factor 0.7

1st Cycle 2nd Cycle 3rd Cycle 4th Cycle 5th Cycle

Figure 5.24: Axial Walking Displacement at Mid-Point with Friction Factor 0.3, 0.5 & 0.7

The axial displacements under the given conditions are plotted according to Tables 5.3 through 5.5, and it is shown as Figure 5.24. It indicates that the axial walking rises as the friction

Number of Operating Cycle (Fig Figure 2.0: Conventional

S-lay installation (Olav Fyrileiv et al., 2005) [9]

Displacement (m)

increases in the range of friction factors. The walking rate per operational cycle with different values of the friction factor is given in Table 5.6:

Table 5.6: Results of Walking Rate

Friction Factor Walking Rate

LB 0.3 Abt. 0.05m/cycle

BE 0.5 Abt. 0.08m/cycle

UB 0.7 Abt. 0.1m/cycle

5.3.2.2 Axial Displacement with Friction Factor 2.0

It can be expected to get the different consequence concerning the walking in accordance with the results of the EAF with friction factor of 2.0, which is discussed in Ch.5.3.1. The Table 5.7 shows the axial displacement at the mid-point of the pipeline with the factor of 2.0:

Table 5.7: Mid-Point Axial Displacement with Friction Factor 2.0

1st Cycle 2nd Cycle 3rd Cycle 4th Cycle 5th Cycle

It can be seen that the axial movement (walking) is reduced compared to the results with other factors (lower values) and gives approximate 0.046m/cycle under the condition. Therefore, the pipelines can partly response like a long pipeline, which is fully restrained over a fixed length under operation conditions, when the axial friction is high enough. As it becomes constrained, the walking phenomenon consequently is to decrease or to disappear.

5.4 Mitigation Measures for Pipeline Walking

This thesis work mainly focuses on pipeline walking which is resulted by cyclic heating and cooling operations of a pipeline especially when the pipeline asymmetrically loaded. This is because the fact that the heating operation is non-uniform, whereas the cooling operation is nearly uniform. [11] Consequently, the instability of pipeline on the seabed occurred by the pipeline walking phenomenon is to be taken into account, and mitigation methods can be discussed by considering the causes of pipeline walking.

This section outlines mitigation measures of pipeline walking based on an understanding of its mechanisms and the results from the case study. The selection of walking mitigation method is evidently depends on the consequential effects of accumulated axial displacement due to pipeline walking. Thus, each mitigation method and its impact on pipeline design are briefly discussed. However, installation, cost of method and planning are not to be presented since it is beyond the scope of this study.

5.4.1 Anchoring

[11]

The pipeline anchors are the most common method to mitigate pipeline walking. It controls and limits the maximum axial displacement by inducing additional tension in the pipeline especially during a shutdown. A typical size of these anchors is in the range of 50 to 350 tons, and its general illustration is shown in Figure 5.25: [1]

Figure 5.25: A Typical Pipeline Restraining Anchor (Ryan Watson et al., 2010) [16]

The pipeline walking phenomenon is associated with the virtual anchor point shifts in the pipeline during heating and cooling operations under asymmetrical load conditions. This means the efficient mitigation can be set by correcting the separation between the virtual anchors. Thus, the end support anchoring can be introduced. It works by way of reducing the separation between the virtual anchors. Anchoring a pipeline especially in the cases of end tension or seabed slope gives equal amount of corrective tension at the opposite end. It functions that the asymmetric

loading is eliminated not to occur walking. Moreover, an anchor can be placed on the pipeline to force the virtual anchor points to share the same location on the pipe.

The tension induced by an anchor can be sufficient for a pipeline to be susceptible to lateral instability (buckling) while an anchor is eliminating walking. In addition, the walking phenomenon can still occur in that buckled region, so walking into a buckle may give over stress into the pipeline. [3] Consequently, it should ensure that the location of an anchor is to be well considered in terms of possible buckling in the pipeline.

5.4.2 Increase Axial Friction

[1]

The walking phenomenon critically depends on the critical axial force. In some cases shows that the walking rate is to decrease due to the growth of an axial friction to a certain extent.

[3] Even the sensitivity study in the thesis work also notes that aspect. Hence, increasing axial friction can be considered to mitigate pipeline walking, and there are several treatments for this achievement.

5.4.2.1 Pipe-Soil Interaction

By investigating a specific pipe-soil interaction on site, it can improve friction factors for design since lower values of friction factors can be expected. However, collecting soil data is quite time consuming activity and costly. Besides, a deep water soil survey is to be one of challenges to get the accurate analysis. It is not just because of the very low shear strength exhibited by most deep water seabed top layers, but also because of the very low effective stress level (a few kPa only). [3] Consequently, it may be impractical to produce such a high axial resistance especially in a deep water pipeline without additional mitigation to control pipeline walking.

5.4.2.2 Pipeline Weight Coating

A concrete weight coating can be a possible way to increase axial and lateral friction resistance which is advantageous to reduce expansion and walking. However, it is likely to experience lateral buckling due to the higher axial resistance, and the higher localized strains may also take place because of higher lateral frictions. [3]

5.4.2.3 Trench and Bury

Those can be used to increase the axial friction resistance in the pipeline. However, a high cost concern and limited equipment for deep water are challenges.

Figure 5.26: A Trenching Operation (from www.theareofdreging.com)

5.4.2.4 Rock Dumping and Mattress

It gives such a function of additional anchoring, so it can reduce the end expansion and walking in addition to buckling controls. However, a large amount of materials is to be required to limit the walking problem. In addition, installation time, cost and recruiting specific vessels for installation are other challenges.

Figure 5.27: A Rock Dumping Operation (from www.nordnes.nl)

5.4.3 Increase Subsea Connection Line Capacity

When it comes to a limit sate for the pipeline, the walking phenomenon is not a failure mode, [2] but it can lead to overstressing on subsea structure connections. Thus, by increasing the capacity of spools and/or jumpers it can accommodate the axial displacement of pipeline induced by walking. However, installation and handling limits may be restrictions for increasing their capacities. [3]

5.5 Summary

This chapter has discussed the results of the FEA on pipeline walking under the thermal cyclic loadings and dealt with different values of the axial friction factors for the sensitivity study. There are no variations in the EAF when the pipeline is fully mobilized (after the 7th load step in this case). From the second heat-up, two VAPs appear and move towards the mid-point and the cold-end respectively as the pipeline gradually heated-up. With regards to pipeline

In document Phenomenon of pipeline walking in high temperature pipeline (sider 52-0)